Kuo-Ping Lin

Revenue forecasting using a least-squares support vector regression model in a fuzzy environment

Revenue forecasting is difficult but essential for companies that want to create high-quality revenue budgets, especially in an uncertain economic environment with changing government policies. Under these conditions, the subjective judgment of decision makers is a crucial factor in making accurate forecasts. This investigation develops a fuzzy least-squares support vector regression model with genetic algorithms (FLSSVRGA) to forecast seasonal revenues. The FLSSVRGA uses the H-level to control the possibility distribution range yielded by the fuzzy model and to provide the fuzzy prediction interval. Depending on various factors, such as the global economy and government policies, a decision maker can elect a different level for H using the FLSSVRGA. The proposed FLSSVRGA model is a rolling forecasting model with time series data updated monthly that predicts revenue for the coming month. Four other forecasting models: the seasonal autoregressive integrated moving average (SARIMA), generalized regression neural networks (GRNN), support vector regression with genetic algorithms (SVRGA) and least-squares support vector regression with genetic algorithms (LSSVRGA), are employed to forecast the same data sets. The experimental results indicate that the FLSSVRGA model outperforms all four models in terms of forecasting accuracy. Thus, the FLSSVRGA model is a useful alternative for forecasting seasonal time series data in an uncertain environment; it can provide a user-defined fuzzy prediction interval for decision makers.

A Comparison of Discrete Algorithms for Fuzzy Weighted Average

Fuzzy weighted average (FWA), which can be applied to various fields such as engineering design, decision analysis, etc., and as function of fuzzy numbers, is suitable for the problem of multiple occurrences of fuzzy parameters. Additional fuzziness may be introduced in the alpha-cut arithmetic. This paper reviews and compares discrete algorithms for the FWAs in both theoretical comparison and numerical comparisons, as opposed to the linear programming algorithms that may be efficient but require the help of linear programming software. An alternative efficient algorithm is also proposed. The algorithm introduces an all-candidate (criteria ratings) weights-replaced benchmark adjusting procedure other than a binary (dichotomy) search in the existing methods. The theoretical worst-case comparison shows that the algorithms of Guu and Guh and our alternative algorithm require the same order of calculation complexity O(n), where n is the number of FWA terms. Yet, the Guh algorithm always requires 2(n-1) calculations to be performed. The algorithm of Guu requires the elemental comparison complexity O(n), and is superior to the other algorithms, in the worst case. On the other hand, our alternative algorithm and Lee and Park algorithm have the same comparison complexity O(n log n). Yet, our alternative algorithm requires O(n) calculation complexity which is better than O(n log n) of the Lee and Park algorithm. The numerical experiments show that the algorithm of Guu and our alternative algorithm generally perform and converge faster by requiring fewer calculations than that of Lee and Parks algorithm. The alternative algorithm requires a slightly smaller number of calculations than that of the Guu algorithm, due to the use of the convergent benchmark adjustment procedure rather than a somewhat fixed binary search. Conversely, in the number of element comparisons, Lee and Parks algorithm is shown numerically generally slightly better than the alternative algorithm due to the simple binary search scheme used. The alternative algorithm also requires fewer comparisons than that of the Guu algorithm. Furthermore, in comparison of the average CPU time requirements, the ranks of the results of these algorithms are consistent with those of the algorithms for the average numbers of calculations. In general, the alternative algorithm may provide an efficient alternative algorithm to the FWAs. In the worse-case situation, the Guu algorithm still may be considered as an efficient alternative

Tourism demand forecasting using novel hybrid system

Accurate prediction of tourism demand is a crucial issue for the tourism and service industry because it can efficiently provide basic information for subsequent tourism planning and policy making. To successfully achieve an accurate prediction of tourism demand, this study develops a novel forecasting system for accurately forecasting tourism demand. The construction of the novel forecasting system combines fuzzy c-means (FCM) with logarithm least-squares support vector regression (LLS-SVR) technologies. Genetic algorithms (GA) were optimally used simultaneously to select the parameters of the LLS-SVR. Data on tourist arrivals to Taiwan and Hong Kong were used. Empirical results indicate that the proposed forecasting system demonstrates a superior performance to other methods in terms of forecasting accuracy.

Time series forecasting by a seasonal support vector regression model

The support vector regression (SVR) model is a novel forecasting approach and has been successfully used to solve time series problems. However, the applications of SVR models in a seasonal time series forecasting has not been widely investigated. This study aims at developing a seasonal support vector regression (SSVR) model to forecast seasonal time series data. Seasonal factors and trends are utilized in the SSVR model to perform forecasts. Furthermore, hybrid genetic algorithms and tabu search (GA/TS) algorithms are applied in order to select three parameters of SSVR models. In this study, two other forecasting models, autoregressive integrated moving average (SARIMA) and SVR are employed for forecasting the same data sets. Empirical results indicate that the SSVR outperforms both SVR and SARIMA models in terms of forecasting accuracy. Thus, the SSVR model is an effective method for seasonal time series forecasting.

Forecasting concentrations of air pollutants by logarithm support vector regression with immune algorithms

Abstract The need to minimize the potential impact of air pollutants on humans has made the accurate prediction of concentrations of air pollutants a crucial subject in environmental research. Support vector regression (SVR) models have been successfully employed to solve time series problems in many fields. The use of SVR models for forecasting concentrations of air pollutants has not been widely investigated. Data preprocessing procedures and the parameter selection of SVR models can radically influence forecasting performance. This study proposes a support vector regression with logarithm preprocessing procedure and immune algorithms (SVRLIA) model which takes advantage of the structural risk minimization of SVR models, the data smoothing of preprocessing procedures, and the optimization of immune algorithms, in order to more accurately forecast concentrations of air pollutants. Three pollutants, namely particulate matter (PM 10 ), nitrogen oxide, (NO x ), and nitrogen dioxide (NO 2 ), are collected and examined to determine the feasibility of the developed SVRLIA model. Experimental results reveal that the SVRLIA model can accurately forecast concentrations of air pollutants.

A fuzzy support vector regression model for business cycle predictions

Business cycle predictions face various sources of uncertainty and imprecision. The uncertainty is usually linguistically determined by the beliefs of decision makers. Thus, the fuzzy set theory is ideally suited to depict vague and uncertain features of business cycle predictions. Consequently, the estimation of fuzzy upper and lower bounds become an essential issue in predicting business cycles in an uncertain environment. The support vector regression (SVR) model is a novel forecasting approach that has been successfully used to solve time series problems. However, the SVR approach has not been widely applied in fuzzy forecasting problems. This study employs support vector regressions to calculate fuzzy upper and lower bounds; and presents a fuzzy support vector regression (FSVR) model for forecasting indices of business cycles. A numerical example of a business cycle prediction in Taiwan was used to demonstrate the forecasting performance of the FSVR model. The empirical results are satisfactory. Therefore, the FSVR model is an effective alternative in forecasting business cycles under uncertain circumstances.

Long-term business cycle forecasting through a potential intuitionistic fuzzy least-squares support vector regression approach

This paper developed a novel intuitionistic fuzzy least-squares support vector regression with genetic algorithms (IFLS-SVRGAs) to accurately forecast the long-term indexes of business cycles. Long-term business cycle forecasting is an important issue in economic evaluation, as business cycle indexes may contain uncertain factors or phenomena such as government policies and financial meltdowns. In order to effectively handle such factors and accidental forecasting indexes of business cycles, the proposed method combined intuitionistic fuzzy technology with least-squares support vector regression (LS-SVR). The LS-SVR method has been successfully applied to forecasting problems, especially time series problems. The prediction model in this paper adopted two LS-SVRs with intuitionistic fuzzy sets, in order to approach the intuitionistic fuzzy upper and lower bounds and to provide numeric prediction values. Furthermore, genetic algorithms (GAs) were simultaneously employed in order to select the parameters of the IFLS-SVR models. In this study, IFLS-SVRGA, intuitionistic fuzzy support vector regression (IFSVR), fuzzy support vector regression (FSVR), least-squares support vector regression (LS-SVR), support vector regression (SVR) and the autoregressive integrated moving average (ARIMA) were employed for the long-term index forecasting of Taiwanese businesses. The empirical results indicated that the proposed IFLS-SVRGA model has better performance in terms of forecasting accuracy than the other methods. Therefore, the IFLS-SVRGA model can efficiently provide credible long-term predictions for business index forecasting in Taiwan.

Using ADABOOST and Rough Set Theory for Predicting Debris Flow Disaster

Debris flow resulting from typhoons, heavy rainfall, tsunamis or other natural disasters is a matter of particular importance to Taiwan owing to the country’s unique geographical environment and exacerbated by poor slope management and global warming. With regard to these types of natural occurrences, recent global events have attracted the attention of experts in various fields, such as civil engineering, environmental engineering and information management. These experts have developed several techniques to study the various factors of debris flow. The ADABOOST and rough set theory (RST) are two emerging methods with regard to classification and rule provision. The ADABOOST, an adaptive boosting machine learning algorithm, uses very little memory during computation and can obtain robust classification results. RST is able to deal with uncertainties and vague information in generating rules for decision makers. Thus, this study develops an ADARST model which uses the unique strengths of the ADABOOST and RST in classification and rule generation and applies the proposed ADARST to analyze debris flow. Specifically, data from previous studies were obtained and used for the purposes of this study. Experimental results have shown that the proposed ADARST model is able to generate better results than those in previous investigations in terms of prediction accuracy. In addition, the designed ADARST model can provide rules including forward and backward reasoning ways for decision makers. Therefore, the proposed ADARST model is shown to be an effective methodology with which to analyze debris flow.

Developing a Fuzzy Bicluster Regression to Estimate Heat Tolerance in Plants by Chlorophyll Fluorescence

This paper presents a straightforward and useful fuzzy regression approach to estimate heat tolerance of plants by chlorophyll fluorescence measurement. The chlorophyll fluorescence measurement is an indicator of functional change of photosynthesis and is sensitive to temperature. Using the fluorescence-temperature curves, the experimenter may determine the heat tolerance (Tc) of plants by intersections of two linear regression lines. However, as traditional statistical regression analysis shows, the experiment may contain uncertain factors or phenomena such as leaf nature and growth environment, which concludes that data may vary among individual plants and different species. This research presents a fuzzy bicluster regression (FBCR) analysis with genetic algorithms, which helps derive a fuzzy intersection set and fuzzy heat tolerance of plants, in addition to the traditional statistical regression analysis. A fuzzy clustering concept and simultaneously optimal determination of data clusters is also developed. Especially, when there are nonlinear inflections in data curves, due to the imperative use of linear regression models, the traditional regression analysis may become unable to sufficiently model the uncertainties exhibited. The FBCR analysis can resolve this problem effectively due to the nonlinear tolerance of the system, even in a linear model. To demonstrate the FBCR analysis, it was applied to estimate the heat tolerance of five plant species. The results derived appeared to be more suitable than that of the conventional method. The approach may provide a useful means for the experimenters to derive more credible results from their chlorophyll fluorescence-temperature data.

Applying Least Squares Support Vector Regression with Genetic Algorithms for Radio-Wave Path-Loss Prediction in Suburban Environment

This paper presents least squares support vector regression with genetic algorithms (LS-SVRGA) models for the prediction of radio-wave path-loss in suburban environment. The least squares support vector regression (LS-SVR) model is a novel forecasting approach and has been successfully used to solve time series problems. However, the application of LS-SVR models in a radio-wave path-loss forecasting has not been widely investigated. This study aims at developing a LS-SVRGA model to forecast radio-wave path-loss data. Furthermore, in the LS-SVRGA model genetic algorithms is applied in order to select two parameters of LS-SVR models. In this study, four forecasting models, Egli, Walfisch and Bertoni (W&B), generalized regression neural networks (GRNN), and support vector regression with genetic algorithms (SVRGA) models are employed for forecasting the same data sets. Empirical results indicate that the LS-SVRGA outperforms others models in terms of forecasting accuracy. Thus, the LS-SVRGA model is an effective method for radio-wave path-loss forecasting in suburban environment.